| --- Log opened Tue Jun 05 00:00:43 2018 |
| 00:21 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has quit [Ping timeout: 121 seconds] |
| 00:33 | | Derakon[AFK] is now known as Derakon |
| 00:53 | | celticminstrel [celticminst@Nightstar-jocbi5.dsl.bell.ca] has joined #code |
| 00:53 | | mode/#code [+o celticminstrel] by ChanServ |
| 00:54 | <@celticminstrel> | Ugh, stupid Apple. |
| 00:54 | <@celticminstrel> | Stop breaking everything! :( |
| 00:55 | < Degi> | What broke |
| 00:55 | <@celticminstrel> | Technically nothing yet but they're deprecating OpenGL apparently. |
| 01:07 | < RchrdB> | This isn't really comforting for games (which really want OpenGL 4) but there probably has to exist a pretty large portion of an OpenGL 3 core profile implementation written atop Metal already, since, as far as I remember, WebGL already works just fine on iOS Safari. |
| 01:08 | <@celticminstrel> | Oh nice. |
| 01:10 | <&McMartin> | Yeah |
| 01:10 | < RchrdB> | Similarly on Windows you have moderately-old-OpenGL implementations atop DirectX version something-or-other shipping in Firefox and Chrome. |
| 01:10 | <&McMartin> | This is super-obnoxious for me though because I don't think SDL1.2 is going to get a Metal backend, which means UQM will stop working *again* right after I've done two emergency releases to deal with software rot |
| 01:11 | <&McMartin> | That *also* said, Windows OpenGL support is top-notch and for a long time led all consumer-grade operating systems. |
| 01:12 | <&McMartin> | https://appleinsider.com/articles/18/06/04/opengl-opencl-deprecated-in-favor-of-metal-2-in-macos-1014-mojave |
| 01:12 | <@celticminstrel> | UQM probably needs updating to SDL2 at some point, right? But I guess that's easier said than done. |
| 01:12 | < RchrdB> | AFAIK one of the things that makes OpenGL a little slower than ideally it ought to be is the fact that libGL.so always gets linked in as a dynamic library, so every call goes through the GOT+PLT or whatever the local equivalent is, which has overhead, and OpenGL is chatty. |
| 01:12 | <&McMartin> | Yeah, SDL2 and SDL1.2 are not only not compatible, you can't sanely switch between them at runtime either because of symbol clashes. |
| 01:13 | <&McMartin> | Good use of modern OpenGL APIs makes OpenGL a *lot* less chatty. |
| 01:13 | <&McMartin> | But people still cling desperately to glVertex4f(). |
| 01:13 | < RchrdB> | I slightly wonder if a source-compatible implementation of OpenGL could be made atop $GRAPHICS_API_OF_THE_WEEK that'd be faster than a real vendor-provided libGL.so just by being able to have inlineable functions and direct calls. |
| 01:13 | <@celticminstrel> | Well I don't think you'd really want to switch between them at runtime... |
| 01:13 | <&McMartin> | It would be during startup, where you were deciding which video/audio backend to use. |
| 01:13 | < RchrdB> | McMartin, I should watch the AZDO presentation for a third time. :) |
| 01:14 | <&McMartin> | AZDO? |
| 01:14 | | Pinkhair [user1@Nightstar-g7hdo5.dyn.optonline.net] has joined #code |
| 01:14 | <@celticminstrel> | Still don't think you'd really want that... |
| 01:14 | <&McMartin> | SDL and SDL2 in fact both do this themselves |
| 01:14 | <@celticminstrel> | ... |
| 01:14 | <@celticminstrel> | "The move was intended to make it easier for developers to port games to Apple's hardware from other platforms, while still taking advantage of technologies like Apple's Metal graphics API." |
| 01:14 | <@celticminstrel> | Um. |
| 01:15 | <@celticminstrel> | Do they really think this is going to have that effect? |
| 01:15 | <&McMartin> | The assembled cultists will eat it up |
| 01:15 | <@celticminstrel> | Isn't it far more likely that developers just drop Mac altogether? |
| 01:15 | <@celticminstrel> | That's what I've been seeing on Twitter, at least. |
| 01:15 | < RchrdB> | Approaching Zero Driver Overhead - it's the documented(ish) this-is-the-really-fast-path-honestly-we-mean-it-guvnor way to use OpenGL 4 + some ubiquitous(ish) extensions to do all kinds of whatever on the GPU with a small fixed number of draw calls per frame. |
| 01:16 | | Pink [user1@Nightstar-g7hdo5.dyn.optonline.net] has quit [Ping timeout: 121 seconds] |
| 01:16 | <&McMartin> | Obviously those people were insufficiently loyal to Glorious Zombie Steve and their works were not worth using anyway |
| 01:16 | <&McMartin> | Aha, neat |
| 01:16 | < RchrdB> | it's a set of ideas wherein you just completely avoid needing more than a few hundred OpenGL function calls |
| 01:16 | | * McMartin is then reminded of the insanity that you can do with mapping memory-mapped I/O into userspace and just not bothering with kernel traps |
| 01:17 | <&McMartin> | How does AZDO compare to, say, Vulkan's lower-level configuration? |
| 01:18 | < RchrdB> | e.g. one of them is that, instead of hundreds of calls to switch which textures are bound as you switch between drawing different models, you use "bindless" textures (an extension which is fairly widely supported, think "pass the pointer to the texture to the shader in a uniform or something instead of having to bind it on the CPU side") or you use a giant (possibly partly-resident) texture atlas |
| 01:20 | < RchrdB> | another is that you use glMultiDrawArraysIndirect(), which is a function that accepts an array of structs, each of which contains the arguments for one glDrawArraysInstancedBaseInstance call, and that *one* call to glMultiDrawArraysIndirect makes the graphics pipeline act like you called glDrawArrays...() a *lot* of times |
| 01:21 | < RchrdB> | (there are a couple of different glMultiDraw* functions) |
| 01:22 | < RchrdB> | McMartin, ehhhhh the way I've seen this described is that in AZDO you're kinda doing slightly-implicitly what with Vulkan you'd just do explicitly |
| 01:24 | < RchrdB> | another one is that instead of making a bunch of function calls to update one GPU buffer at a time, you call glBufferStorage() in just the right way, get back a void* on the host side which happens to be mapped to actual real GPU memory, and then you can write directly into memory that can be accessed from the GPU side without the overhead of calling into the driver for each write |
| 01:26 | < RchrdB> | https://www.youtube.com/watch?v=K70QbvzB6II |
| 01:27 | < RchrdB> | (was hosted somewhere on GDC's website for a while but hey someone appears to have shoved it on youtube too recently) |
| 01:35 | | Kindamoody is now known as Kindamoody[zZz] |
| 01:40 | | Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 01:45 | | * McMartin would like to see Vulkan much more broadly deployed and then OpenGL moved entirely into userspace, talking to Vulkan |
| 01:45 | <&McMartin> | I wonder how much that would end up costing |
| 01:45 | <&McMartin> | But that would pull *so much shit* out of the kernel, holy cheese |
| 02:13 | | RchrdB [RchrdB@Nightstar-qe9.aug.187.81.IP] has quit [Connection closed] |
| 03:34 | | Degi [Degi@Nightstar-f2tlg7.dyn.telefonica.de] has quit [Connection closed] |
| 04:06 | < Mahal> | https://docs.google.com/presentation/d/17bKudNDduvN-7hWv7S84MiHUj2AnOPNbwjTM8euDC8w/mobilepresent#slide=id.g3b3c426762_0_228 interesting rant/talk on the topid of security, moore's law, and the anomaly of cheap complexity. |
| 04:09 | <&McMartin> | "How did this research code someone wrote in two weeks 20 years ago end up in a billion devices?" |
| 04:18 | <&McMartin> | I do find myself wondering how much of those issues persist with extremely weak microcontrollers. Some do, but... |
| 04:18 | <&McMartin> | PIC, Mask ROM, no writable nonvolatile storage. |
| 04:19 | <&McMartin> | (That said, oh god, to Hell with the PIC) |
| 04:29 | | Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has joined #code |
| 04:29 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 04:55 | | celticminstrel [celticminst@Nightstar-jocbi5.dsl.bell.ca] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
| 05:33 | | Derakon is now known as Derakon[AFK] |
| 07:03 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has joined #code |
| 07:03 | | mode/#code [+o himi] by ChanServ |
| 07:11 | | Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 07:18 | | Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has joined #code |
| 07:18 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 07:23 | | Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 08:33 | | Kindamoody[zZz] is now known as Kindamoody |
| 08:40 | | Vorntastic [uid293981@Nightstar-6br85t.irccloud.com] has joined #code |
| 08:50 | | mac [macdjord@Nightstar-ahbhn1.cable.rogers.com] has quit [Ping timeout: 121 seconds] |
| 08:53 | | macdjord [macdjord@Nightstar-ahbhn1.cable.rogers.com] has joined #code |
| 08:53 | | mode/#code [+o macdjord] by ChanServ |
| 09:04 | | Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has joined #code |
| 10:10 | | gnolam_ [lenin@Nightstar-ego6cb.cust.bahnhof.se] has joined #code |
| 10:14 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has quit [Ping timeout: 121 seconds] |
| 10:54 | | majo [NSkiwiirc@Nightstar-9cmf82.se.alltele.net] has joined #code |
| 11:59 | | Degi [Degi@Nightstar-f2tlg7.dyn.telefonica.de] has joined #code |
| 12:00 | | Jessikat [Jessikat@Nightstar-av3.tcr.132.82.IP] has joined #code |
| 12:16 | | majo [NSkiwiirc@Nightstar-9cmf82.se.alltele.net] has quit [[NS] Quit: http://www.kiwiirc.com/ - A hand crafted IRC client] |
| 12:29 | | Kindamoody is now known as Kindamoody|afk |
| 12:31 | | Emmy is now known as Emmy-afk |
| 12:36 | | Jessikat` [Jessikat@Nightstar-av3.tcr.132.82.IP] has joined #code |
| 12:36 | | Jessikat [Jessikat@Nightstar-av3.tcr.132.82.IP] has quit [The TLS connection was non-properly terminated.] |
| 12:38 | | * TheWatcher vaguely ponders spinning up his own gitlab instance |
| 13:57 | | Jessikat [Jessikat@Nightstar-3adkl9.dab.02.net] has joined #code |
| 13:59 | | Jessikat` [Jessikat@Nightstar-av3.tcr.132.82.IP] has quit [Ping timeout: 121 seconds] |
| 14:22 | | Jessikat` [Jessikat@Nightstar-3adkl9.dab.02.net] has joined #code |
| 14:22 | | Jessikat [Jessikat@Nightstar-3adkl9.dab.02.net] has quit [The TLS connection was non-properly terminated.] |
| 15:06 | | Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has joined #code |
| 15:07 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 15:11 | <&McMartin> | TheWatcher: Given your endless travails with, e.g., MediaWiki and That Coursework Thing it would be interesting to hear how Gitlab ends up comparing. |
| 15:12 | <@TheWatcher> | Well, I run GitLab in work, too - I'm just meaning for myself on one of my home servers. |
| 15:15 | <@TheWatcher> | Gitlab is.... okay. If you're running a distro the omnibus edition supports (ubuntu, debian, centos, opensuse), it's simple to get going. From source it's rather more of a pain simply from a dependencies point of view, especially as there have been at least two occasions where the docs have said - for example - one version of ruby when what they /actually/ meant was a |
| 15:15 | <@TheWatcher> | different version. |
| 15:17 | <&McMartin> | I'd say "hooray, gg", but I've been responsible for maintaining build instructions before and it is not a task suitable for humans :( |
| 15:17 | <@TheWatcher> | It's also steadily accreting languages as it evolves |
| 15:17 | <@TheWatcher> | which makes that whole 'about those dependencies...' thing a big more of aheadache as time goes on. |
| 15:18 | <~Vornicus> | fucking hate dependency stuff |
| 15:19 | <@TheWatcher> | But once you have it up and working, it is good, and certainly in work it has been easily handling constant daily use by over thousand users with barely any issues. |
| 15:20 | <&ToxicFrog> | I've considered it as well. NixOS has a services.gitlab namespace, so it probably wouldn't be at all hard to power up, but that doesn't communicate anything about how hard it is to keep it happy once it's running. |
| 15:20 | <&ToxicFrog> | At the moment I'm just using gitolite. |
| 15:23 | <@TheWatcher> | My chief issues with gitlab boil down to the kludges (and make no mistake, that's what they are) they've implemented to make https access to repos work, and the fact that they have a monthly release schedule that exactly coincides with the start of semester in September >.> |
| 15:24 | <@TheWatcher> | https://docs.gitlab.com/ee/development/architecture.html#components - it's not exactly clean |
| 15:25 | <&ToxicFrog> | D: |
| 15:29 | <@TheWatcher> | With bits of that stuff being done in ruby, some in python, workhorse is go, and node used in the background. |
| 15:32 | <~Vornicus> | whyyyyyyyy |
| 15:33 | <&ToxicFrog> | This leads me somewhat uncomfortably to the realization that as a user and developer, I want lots of small, highly specialized tools I can easily compose |
| 15:33 | <&ToxicFrog> | But as a system administrator, I want one big monolithic thing I can turn on and forget about |
| 15:36 | <@TheWatcher> | Vornicus: I think part of it is just down to the fact that it has evolved rather than been designed, and along the way things have cropped up that they weren't expecting, and they've bolted on things to make it work |
| 15:44 | <@TheWatcher> | Like the workhorse thing. That's needed because originally https git access was going User -> (apache|nginx) -> Unicorn (and hence the rails app) which was doing direct interaction with the repos. Which worked, for small repos and small pushes/pulls, but was running into timeout problems for larger stuff. So one of the devs, who had been learing go, decided to try and make |
| 15:44 | <@TheWatcher> | something to sit between unicorn and the web server to handle git operations while passing through other stuff. |
| 15:49 | | Vorntastic [uid293981@Nightstar-6br85t.irccloud.com] has quit [[NS] Quit: Connection closed for inactivity] |
| 16:09 | | Jessikat` is now known as Jessikat |
| 16:11 | | Jessikat` [Jessikat@Nightstar-3adkl9.dab.02.net] has joined #code |
| 16:11 | | Jessikat [Jessikat@Nightstar-3adkl9.dab.02.net] has quit [The TLS connection was non-properly terminated.] |
| 16:13 | | Jessikat [Jessikat@Nightstar-3adkl9.dab.02.net] has joined #code |
| 16:15 | | Jessikat` [Jessikat@Nightstar-3adkl9.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 16:43 | | Vorntastic [uid293981@Nightstar-6br85t.irccloud.com] has joined #code |
| 16:45 | <&[R]> | I got gitlab working on Slackware, don't recall having any issues, beyond discovering the Ruby's upstream version numbers are brain-dead. |
| 16:46 | <&[R]> | (WTF does upstream need revision numbers to patches?) |
| 16:58 | <@TheWatcher> | I'm not commenting on ruby and brain-deadness |
| 16:58 | <@abudhabi> | >tfw when you come back from a short trip to the countryside to see that the website is down for no discernible reason |
| 16:59 | <@TheWatcher> | I guess, though, I should be grateful that they wrote the core of gitlab in ruby, and didn't decide to use node.... |
| 16:59 | <@TheWatcher> | abudhabi: nowt in the logs? :( |
| 17:02 | <@abudhabi> | TheWatcher: The *mail* server is up and running. But HTTP(S) just goes all 'connection refused'. |
| 17:03 | <@abudhabi> | The last thing I did after going away was installing roundcube from the repository. |
| 17:03 | <@abudhabi> | It also installed php5 again (after I've upgraded to 7). php -v still says 7 is running. |
| 17:03 | <@abudhabi> | But I doubt that would be causing the connection refused. |
| 17:04 | <@TheWatcher> | `ps aux | grep apache` showing workers? |
| 17:04 | <&[R]> | The php-cli and mod_php can run entirely different versions |
| 17:05 | <&[R]> | Testing with the cli to see the version is completely useless unless you're actually using it for stuff. |
| 17:05 | <&[R]> | Load up a phpinfo page instead |
| 17:05 | <@abudhabi> | TheWatcher: That yields one line. |
| 17:05 | <@abudhabi> | [R]: I would, except CONNECTION_REFUSED. |
| 17:05 | <@TheWatcher> | abudhabi: " grep --colour=auto apache" I bet |
| 17:05 | <@TheWatcher> | In which case, apache fallen over. |
| 17:06 | <@abudhabi> | Huh. |
| 17:06 | <&[R]> | Right, but that's not an issue with PHP. That's something Apache would be failing at. |
| 17:06 | <&[R]> | If it were PHP it'd be an Internal Server Error |
| 17:06 | <@abudhabi> | Tried restarting apache2. |
| 17:06 | <@abudhabi> | "Segmentation fault (core dumped)" |
| 17:06 | <@abudhabi> | Welp. |
| 17:08 | <@TheWatcher> | Anything in /var/log/apache/error_log ? |
| 17:09 | <@TheWatcher> | *apache2/error_log |
| 17:10 | <@abudhabi> | Didn't check, simply undid the last thing I did - removed roundcube. This removed also all the php5 bits, and apache2 restarted properly. |
| 17:10 | | * TheWatcher nod |
| 17:10 | <@abudhabi> | Thanks for helping me troubleshoot, guys. |
| 17:11 | <@TheWatcher> | Sure, no prob |
| 17:39 | | Emmy-afk is now known as Emmy |
| --- Log closed Tue Jun 05 18:12:39 2018 |
| --- Log opened Tue Jun 05 18:12:47 2018 |
| 18:12 | | TheWatcher [chris@GlobalOperator.Nightstar.Net] has joined #code |
| 18:12 | | Irssi: #code: Total of 37 nicks [25 ops, 0 halfops, 0 voices, 12 normal] |
| 18:12 | | mode/#code [+o TheWatcher] by ChanServ |
| 18:12 | | Irssi: Join to #code was synced in 13 secs |
| 18:49 | < Vorntastic> | Huh |
| 18:50 | < Vorntastic> | I didn't realize it until now but clock speeds have really not moved much lately |
| 18:53 | <@TheWatcher> | Nope |
| 18:53 | <&McMartin> | Feels like a decade, maybe more. |
| 18:53 | | gnolam_ is now known as gnolam |
| 18:54 | | mode/#code [+o gnolam] by ChanServ |
| 18:56 | < Vorntastic> | Intel is apparently just now getting 5ghz |
| 18:56 | <@abudhabi> | I should perhaps invest in some CPU that's got high hertz, but not too many cores. The kinds of games I play tend not to use many cores very well. |
| 18:57 | < Vorntastic> | I remember my 2003 Mac getting 1g, and my 1991 mac getting 20m |
| 18:58 | < Vorntastic> | So in 15 years it went like 5 fold and the 12 before that, 50 |
| 19:02 | <@TheWatcher> | http://preshing.com/20120208/a-look-back-at-single-threaded-cpu-performance/ is a decent overview, but yeah, we've sorta plateaued on CPU speed, and even parallelism isn't helping much for general computing now. |
| 19:04 | <@TheWatcher> | that article's old, and since then it's levelled-off even more |
| 19:07 | <@TheWatcher> | There's heat problems with higher clock speeds and we're right up against the limit of how small we can go without quantum effects screwing up gate behaviour |
| 19:10 | <@TheWatcher> | (ISTR that Berkeley somehow managed to get a 1nm transistor to work, so maybe we can get around electron tunnelling, but I've not heard of anything using the tech in the wild) |
| 19:16 | < Jessikat> | Games are fundamentally bad at using many cores |
| 19:16 | < Jessikat> | Unless they do a lot of render or fancy physics kind of work that parallelises easily |
| 19:22 | <@gnolam> | TheWatcher: and then we still have the problem that power consumption increases with the square of the clock frequency. |
| 19:23 | | Vornotron [Vorn@Nightstar-sjaki9.res.rr.com] has joined #code |
| 19:23 | | mode/#code [+qo Vornotron Vornotron] by ChanServ |
| 19:26 | | Vornicus [Vorn@Nightstar-sjaki9.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 20:03 | | Jessikat [Jessikat@Nightstar-3adkl9.dab.02.net] has quit [[NS] Quit: Bye] |
| 21:14 | | Kindamoody|afk is now known as Kindamoody |
| 22:11 | | Degi_ [Degi@Nightstar-otqivl.dyn.telefonica.de] has joined #code |
| 22:14 | | Degi [Degi@Nightstar-f2tlg7.dyn.telefonica.de] has quit [Ping timeout: 121 seconds] |
| 22:14 | <@TheWatcher> | gnolam: true :/ |
| 22:17 | | * abudhabi cargo-cult admins onwards, into the land of actually installing roundcube without forcing apache2 to commit seppuku. |
| 22:18 | <@gnolam> | It's at least possible there's a technical fix for it though. |
| 22:18 | | Kindamoody is now known as Kindamoody[zZz] |
| 22:19 | <@gnolam> | Since one of the terms that makes it O(n^2) is that you basically have to increase the transistor voltage in step with the frequency. |
| 22:21 | <@TheWatcher> | abudhabi: installing it from source? |
| 22:22 | <@abudhabi> | TheWatcher: Using the installer. |
| 22:22 | <@abudhabi> | Presently trying to figure out how to install Net_LDAP3 which is in some third-party repo. |
| 22:23 | <@TheWatcher> | gnolam: on looking, the berkeley approach uses a different transistor tech based on carbon nanotubes and MoS2 |
| 22:30 | <@abudhabi> | TheWatcher: Any idea how to install this thing https://git.kolab.org/diffusion/PNL/ ? |
| 22:31 | <@abudhabi> | I can certainly download it, but where do I put it? |
| 22:33 | <@TheWatcher> | you're on ubuntu? Doesn't `sudo apt-get install php-net-ldap3` work? |
| 22:34 | <@abudhabi> | Negative. |
| 22:34 | <@abudhabi> | E: Unable to locate package php-net-ldap3 |
| 22:34 | | macdjord [macdjord@Nightstar-ahbhn1.cable.rogers.com] has quit [[NS] Quit: Pie Jesu Domine (*thunk*), Donna Eis Requiem (*thunk*).] |
| 22:35 | | Degi_ is now known as Degi |
| 22:38 | <@abudhabi> | I think I'm going to disregard it, since it's not strictly required. |
| 22:39 | <@TheWatcher> | abudhabi: in theory, it you stick the contents of the lib/ directory into (possibly) /usr/share/php/ it might work |
| 22:40 | <@TheWatcher> | so you'd have /usr/share/php/LDAP3.php and /usr/share/php/LDAP3/ ... |
| 22:40 | <@TheWatcher> | it should really be installable through pear, but it doesn't seem to be in the standard channels |
| 22:43 | | Vorntastic [uid293981@Nightstar-6br85t.irccloud.com] has quit [[NS] Quit: Connection closed for inactivity] |
| 22:44 | <@abudhabi> | TheWatcher: That worked! |
| 22:46 | <@TheWatcher> | Goodo. |
| --- Log closed Tue Jun 05 22:46:31 2018 |
| --- Log opened Tue Jun 05 22:47:00 2018 |
| 22:47 | | TheWatcher [chris@GlobalOperator.Nightstar.Net] has joined #code |
| 22:47 | | Irssi: #code: Total of 34 nicks [25 ops, 0 halfops, 0 voices, 9 normal] |
| 22:47 | | mode/#code [+o TheWatcher] by ChanServ |
| 22:47 | | Irssi: Join to #code was synced in 13 secs |
| 22:53 | | Emmy [Emmy@Nightstar-9p7hb1.direct-adsl.nl] has quit [Ping timeout: 121 seconds] |
| 23:13 | | Vornotron [Vorn@Nightstar-sjaki9.res.rr.com] has quit [Ping timeout: 121 seconds] |
| --- Log closed Wed Jun 06 00:00:55 2018 |